This week’s tech selloff shows Wall Street is getting nervous about businesses betting the bank on AI

Probably. But over the past week, investors made clear that they think it will happen later rather than sooner.

During the first half of this year, Wall Street sent tech shares—and those of many other businesses, for that matter—soaring on the hope that the hundreds of billions of dollars they’ve spent so far on AI would pay off quickly. Companies that played the AI game right would soon enjoy a windfall of cost savings and extra profits, or so the theory went.

But then a funny thing happened this week: The Nasdaq nosedived as much as 5% after Wall Street started worrying that the expected AI utopia may not materialize as quickly as initially believed (tech shares recovered somewhat in mid-day trading on Friday). You see, most companies betting the bank on AI have little financial benefit to show from it so far.

Alphabet CEO Sundar Pichai was unusually candid about the current spend-now-ask-questions-later mentality permeating the tech sector. In a quarterly earnings call with analysts on Tuesday, he brushed off any criticism of his company’s ballooning AI bill, saying, “When we go through a curve like this, the risk of under-investing is dramatically greater than the risk of over-investing for us here.”

Investors cringed. Despite Alphabet’s better-than-expected earnings, its shares, as of mid-day Friday, had fallen 6% this week.

Next week, Meta, Microsoft, Amazon, and Apple will also be in the firing line over earnings. It’s a good bet Mark Zuckerberg, Satya Nadella, Andy Jassy, and Tim Cook will also have to defend their big spending on everything AI.

Don’t get me wrong—I’m no Debbie Downer about AI. It will be hugely transformative for business, from handling back-office chores to taking notes for doctors when they see patients. But I’m also a realist. Most companies are still in the very early stages of deploying the technology and wider rollouts will take time, especially considering the very real risks of hallucinations and inaccuracies. Most of the promised financial benefits—the workforce synergies, to cite one corporate euphemism—will likely come much further down the road.

Sure, there are a handful of exceptions, mostly involving companies selling the basic tools that customers need to deploy AI. Those obvious winners include chipmaker Nvidia and big cloud service providers Alphabet, Microsoft, and Amazon, although they’re also having to invest huge amounts of money to expand their data centers to keep up with demand.

So is Wall Street right to be nervous about its initial bet on a quick payoff for AI? Yes. As history shows, tech is indeed transformative, but over the longer term. Online retailing, for example, took years to become viable after the Dotcom crash, wiping out many investors in the process. Meanwhile, video streaming and social media also took their sweet time due to initially slow consumer interest in watching movies online and glitchy technology for connecting with friends. AI is no different in that it won’t necessarily adhere to Wall Street’s timeline.

Verne Kopytoff

Want to send thoughts or suggestions to Data Sheet? Drop a line here.

Today’s edition of Data Sheet was curated by David Meyer.

NEWSWORTHY

SpaceX Falcon 9 clearance. The Federal Aviation Administration has cleared SpaceX’s Falcon 9 rocket for a return to action, Reuters reports. The Falcon 9 had been grounded since a failed launch a couple of weeks back—the workhorse rocket’s first in over seven years. Elon Musk’s company says it could return the Falcon 9 to service as soon as tomorrow.

No Fortnite for Samsung. Epic Games has yanked Fortnite from Samsung’s Galaxy Store, partly in protest at the smartphone maker’s decision to no longer allow the sideloading—i.e. installation not via an app store—of apps. As Epic explained in a blog post, the other reason was “public revelations in the U.S. Epic v Google lawsuit of ongoing Google proposals to Samsung to restrain competition in the market for Android app distribution.”

Runway training data scandal. 404 Media reports that Runway, the hot AI video generation service, trained its latest model on thousands of YouTube videos and pirated films. Noting that Runway’s backers include Google (and therefore YouTube) parent Alphabet, The Verge asked Google for comment, only to be pointed to a previous statement saying such training is a “clear violation” of YouTube’s policies.

ON OUR FEED

“The United States currently has a lead in AI development, but continued leadership is far from guaranteed. Authoritarian governments the world over are willing to spend enormous amounts of money to catch up and ultimately overtake us.”

—OpenAI CEO Sam Altman uses a Washington Post op-ed to claim that AI control is “the urgent question of our time” and to call for “a U.S.-led global coalition of like-minded countries and an innovative new strategy” to ensure that AI benefits “the most people possible.”

IN CASE YOU MISSED IT

OpenAI’s new SearchGPT takes aim at Google in the battle for AI search dominance. But will it win the war?, by Sharon Goldman

Startup with ‘radical’ concept for AI chips emerges from stealth with $15 million to try to challenge Nvidia, by Jeremy Kahn

Meta’s WhatsApp cracks 100 million as the U.S. discovers the bridge between ‘blue and green’ bubbles, by Bloomberg

A post about Kamala Harris pushed a long-simmering feud between tech execs into the open, by Paolo Confino

LGBTQ+ dating app aims to protect closeted athletes by disabling features at Olympic facilities, by Marco Quiroz-Gutierrez

Jack Dorsey is about to overhaul Block in a reorg he warns may feel ‘big and disruptive or uncomfortable’—internal memo, by Kali Hays

Elon Musk’s transgender daughter accuses ‘narcissistic’ entrepreneur of lying about her childhood since he was never there, by Orianna Rosa Royle

BEFORE YOU GO

Ferrari deepfake scam attempt. Bloomberg has a fascinating story about how sophisticated scammers tried to get to a Ferrari executive. First, they sent the executive WhatsApp messages purporting to come from CEO Benedetto Vigna, then they followed up with a phone call from Vigna’s deepfaked voice. Whatever the aim of the exercise, it ended abruptly after the unnamed exec asked the voice a question to which only Vigna would know the answer.